JP2018169978A - Control apparatus, control system, control method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To contribute to improving transmission speed between a host machine and a guest machine.SOLUTION: A control apparatus includes one or more virtual machines controlling virtual network function, and a host machine controlling the virtual machines. The control apparatus includes: a virtual control unit for controlling the virtual machines; a physical control unit for controlling the host machine; and a first storage unit including a first storage area and a second storage area, and accessible from the virtual control unit and the physical control unit. The first storage area stores packet data. The second storage area stores a virtual address specifying the first storage area. The control apparatus specifies the first storage area on the basis of the virtual address stored in the second storage area, to read the packet data from the first storage area.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a control device, a control system, a control method, and a program.

  The use of network function virtualization (NFV), which virtualizes functions of network devices (routers, gateways, firewalls, load balancers, etc.), is spreading. In NFV, the virtual machine substitutes for the function of the network device by executing the function of the network device on the virtual machine. In recent years, NFV has been required to improve network performance.

  In Patent Document 1, a storage relationship between a virtual machine and a computer node is searched based on a computer resource amount for each computer node of a physical infrastructure and a network resource amount for each link, and a virtual machine and a virtual connection are found on the physical infrastructure. The technology to build is described.

  Japanese Patent Application Laid-Open No. 2004-228688 describes a technique for constructing a virtual communication path that connects an external process operable in a host OS and a virtual machine in a guest OS (Operating System) operating in the virtual machine.

  Patent Document 3 describes a technique in which network functions are arranged in one or more virtual containers, and virtual containers are grouped using group identifiers corresponding to virtual network nodes. In the technique described in Patent Document 3, even when a network function is distributed in a plurality of virtual containers, grouping of virtual containers facilitates resource allocation to each virtual container. .

  Here, Non-Patent Document 1 discloses a configuration for realizing NFV as an ETSI (European Telecommunications Standards Institute) NFV standard.

  Furthermore, in recent years, some of NFVI (Network Function Virtualization Infrastructure) and VNF (Virtual Network Function) are part of ETSI NFV for the purpose of improving network performance and reducing total cost of ownership (TCO). It is considered to offload processing using an accelerator. For example, using an accelerator such as a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), or a network processing unit (NPU) to offload a part of processing of NFVI and VNF is being studied.

  Non-Patent Document 2 describes the effect on the functional components that make up the NFV and the effect on the reference points between the functional components by applying an accelerator in ETSI NFV. Has been.

  Non-Patent Document 3 (particularly “5 Use Cases”) describes an example of a VNF use case in which some processing of NFVI and VNF is offloaded using an accelerator in ETSI NFV.

JP, 2015-162147, A JP2015-197874A JP-T-2006-509212

“Network Functions Virtualization (NFV); Architectural Framework”, ETSI GS NFV 002 V1.1.1 (2013-10), [online], search on February 16, 2017, Internet <URL: http: / /www.etsi.org/deliver/etsi_gs/nfv/001_099/002/01.01.01_60/gs_nfv002v010101p.pdf> “Network Functions Virtualization (NFV); Acceleration Technologies; Management Aspects Specification”, ETSI GS NFV-IFA 004 V2.1. <URL: http://www.etsi.org/deliver/etsi_gs/NFV-IFA/001_099/004/02.01.01_60/gs_NFV-IFA004v020101p.pdf> “Network Functions Virtualization (NFV); Acceleration Technologies; Report on Acceleration Technologies & Use Cases”, ETSI GS NFV-IFA 001, 29, 2009, 2020. Day search, Internet <URL: http://www.etsi.org/deliver/etsi_gs/NFV-IFA/001_099/001/01.01.01_60/gs_nfv-ifa001v010101p.pdf>

  The disclosure of the above prior art document is incorporated herein by reference. The following analysis has been made from the viewpoint of the present invention.

  As described above, Non-Patent Document 2 describes the effect of applying an accelerator to functional components that constitute an NFV. Non-Patent Document 3 describes an example of a VNF use case in which an accelerator is used to offload a part of processing of NFVI and VNF. However, Non-Patent Documents 1 to 3 do not describe improvement in communication speed between the host machine and the guest machine in the configuration to which the accelerator is applied.

  In addition, in the techniques described in Patent Documents 1 to 3, there is no description of a method for improving communication speed between a host machine and a guest machine using an accelerator in NFV.

  SUMMARY An advantage of some aspects of the invention is that it provides a control device, a control system, a control method, and a program that contribute to an improvement in communication speed between a host machine and a guest machine.

According to a first aspect of the present invention, a control device is provided. The control device includes one or more virtual machines that control the virtual network function and a host machine that controls the virtual machines.
The control device includes a virtual control unit that controls the virtual machine.
Further, the control device includes a physical control unit that controls the host machine.
Furthermore, the control device includes a first storage unit that is accessible to the virtual control unit and the physical control unit.
The first storage unit includes a first storage area and a second storage area.
In the first storage area, packet data is stored.
The second storage area stores a virtual address that identifies the first storage area.
Further, the control device specifies the first storage area based on the virtual address stored in the second storage area, and reads packet data from the specified first storage area.

According to a second aspect of the present invention, a control system is provided. The control system includes one or more virtual machines that control a virtual network function, a control device that includes a host machine that controls the virtual machines, and a management node that allocates the virtual machines. Composed.
The control device includes a virtual control unit that controls the virtual machine.
Further, the control device includes a physical control unit that controls the host machine.
The control device further includes a first storage unit that is accessible to the virtual control unit and the physical control unit, and the first storage unit includes a first storage area and a second storage area. It is comprised including.
In the first storage area, packet data is stored.
The second storage area stores a virtual address that identifies the first storage area.
Further, the control device specifies the first storage area based on the virtual address stored in the second storage area, and reads packet data from the specified first storage area.

According to a third aspect of the present invention, a control method is provided. The control method is a control method of the control device. The control device includes one or more virtual machines that control a virtual network function and a host machine that controls the virtual machines, and controls a virtual control unit that controls the virtual machines, and the host machine. A physical control unit that controls, a first storage unit that is accessible to the virtual control unit and the physical control unit, wherein the first storage unit includes a first storage area, a second storage unit, and a second storage unit. And a storage area.
The control method includes a step of storing packet data in the first storage area.
The control method includes a step of storing a virtual address that identifies the first storage area in the second storage area.
The control method includes a step of specifying the first storage area based on the virtual address stored in the second storage area.
The control method includes a step of reading packet data from the identified first storage area.
The present method is linked to a specific machine called a control device including one or more virtual machines that control the virtual network function and a host machine that controls the virtual machines.

According to a fourth aspect of the present invention, a program is provided. The program is a program that is executed by a computer that controls the control device. The control device includes one or more virtual machines that control a virtual network function and a host machine that controls the virtual machines, and controls a virtual control unit that controls the virtual machines, and the host machine. A physical control unit that controls, a first storage unit that is accessible to the virtual control unit and the physical control unit, wherein the first storage unit includes a first storage area, a second storage unit, and a second storage unit. And a storage area.
The program causes the computer to execute a process of storing packet data in the first storage area.
The program causes the computer to execute a process of storing a virtual address specifying the first storage area in the second storage area.
The program causes the program to execute a process of specifying the first storage area based on the virtual address stored in the second storage area.
The program causes the program to execute a process of reading packet data from the identified first storage area.
The program can be recorded on a computer-readable storage medium. The storage medium may be non-transient such as a semiconductor memory, a hard disk, a magnetic recording medium, an optical recording medium, or the like. The present invention can also be embodied as a computer program product.

  According to each aspect of the present invention, a control device, a control system, a control method, and a program that contribute to an improvement in communication speed between a host machine and a guest machine are provided.

It is a figure for demonstrating the outline | summary of one Embodiment. 2 is a block diagram illustrating an example of an internal configuration of a communication system 100. FIG. It is a block diagram which shows an example of an internal structure of NFVI10. It is a sequence diagram which shows an example of operation | movement of the communication system 100 which concerns on 1st Embodiment. It is a sequence diagram which shows an example of operation | movement of the communication system 100 which concerns on 1st Embodiment. It is a sequence diagram which shows an example of operation | movement of the communication system 100 which concerns on 1st Embodiment. It is a sequence diagram which shows an example of operation | movement of the communication system 100 which concerns on 2nd Embodiment. It is a sequence diagram which shows an example of operation | movement of the communication system 100 which concerns on 2nd Embodiment. It is a sequence diagram which shows an example of operation | movement of the communication system 100 which concerns on 2nd Embodiment.

  First, an outline of an embodiment will be described with reference to FIG. Note that the reference numerals of the drawings attached to the outline are attached to the respective elements for convenience as an example for facilitating understanding, and the description of the outline is not intended to be any limitation. The connection lines between the blocks in each block diagram include both bidirectional and unidirectional. The unidirectional arrow schematically shows the main signal (data) flow and does not exclude bidirectionality.

  As described above, a control device that contributes to improving the communication speed between the host machine and the guest machine is desired.

  Therefore, as an example, the control apparatus 1000 shown in FIG. 1 is provided. The control device 1000 includes one or more virtual machines that control the virtual network function and a host machine that controls the virtual machines. The control device 1000 includes a virtual control unit 1001, a physical control unit 1002, and a first storage unit 1003.

  The virtual control unit 1001 controls the virtual machine. The physical control unit 1002 controls the host machine.

  The first storage unit 1003 includes a first storage area 1004 and a second storage area 1005, and is accessible to the virtual control unit 1001 and the physical control unit 1002. The first storage area 1004 stores packet data. The second storage area 1005 stores a virtual address that identifies the first storage area 1004.

  The control device 1000 identifies the first storage area 1004 based on the virtual address stored in the second storage area 1005, and reads the packet data from the identified first storage area 1004.

  Here, the control device 1000 stores the packet data in a storage area (first storage area 1004) accessible from both the host machine and the guest machine. For this reason, when transmitting and receiving packet data between the host machine and the guest machine, the control apparatus 1000 does not need to copy the packet data from the transmission source to the transmission destination. Therefore, the control apparatus 1000 contributes to improving the communication speed between the host machine and the guest machine.

[First Embodiment]
The first embodiment will be described in more detail with reference to the drawings.

  FIG. 2 is a block diagram illustrating an example of an internal configuration of the communication system 100 according to the present embodiment.

  Referring to FIG. 2, the communication system 100 according to the present embodiment includes an NFV MANO (Network Function Virtualization Management and Orchestration) 2, a VNF layer 3, and an NFVI 10. The NFV MANO2, the VNF layer 3, and the NFVI 10 communicate with the OSS / BSS4. 2 is an example of the communication system 100 according to the present embodiment, and the communication system 100 according to the present embodiment is not intended to be limited to the configuration illustrated in FIG.

  First, NFV MANO2 will be described in detail.

  The NFV MANO 2 includes a VNF orchestrator 21, one or more VNF managers 22, and a VIM (Virtualized Infrastructure Manager) 23.

  The VNF orchestrator 21 receives an instruction for generating a VNF from the client 200.

  The VNF manager 22 manages the life cycle of the VNF. The VNF manager 22 may be arranged in the NFV MANO 2 for each VNF. Alternatively, one VNF manager 22 may manage the life cycle of two or more VNFs.

  The VIM 23 executes a process for managing VNF resources.

(VNF layer 3)
Next, the VNF layer 3 will be described in detail.

  The VNF layer 3 includes virtual nodes (VNF) (31-a to 31-c) and management virtual nodes (EMS (Element Management System)) (32-a to 32-c). In the following description, the virtual nodes (VNF) (31-a to 31-c) and the management virtual nodes (EMS) (32-a to 32-c) need not be distinguished from each other. They are represented as a virtual node (VNF) 31 and a management virtual node (EMS) 32. FIG. 2 also shows three virtual nodes (VNF) (31-a to 31-c) and three management virtual nodes (EMS) (32-a to 32-c). VNF) 31 and management virtual node (EMS) 32 are not limited to three.

  By setting (implementing) a network function in the virtual node (VNF) 31, the virtual node (VNF) 31 realizes VNF. The virtual machine executes the network function set in the virtual node (VNF) 31 using the virtual node (VNF) 31 in which the network function is set. Of course, one virtual node (VNF) 31 may correspond to two or more virtual machines. Of course, one virtual machine may use two or more virtual nodes (VNFs) 31.

  For example, the virtual node (VNF) 31 is an application program based on Intel (registered trademark) DPDK (Data Plane Development Kit) (hereinafter simply referred to as an application). For example, applications targeted for VNF include IP Sec tunnels Termination, Virtual Base Station (VBS) L1, Transcoding, NFVI Virtual Network, etc., described in “Use Case” of Chapter 5 of Non-Patent Document 3. including. Further, applications targeted for VNF include vCPE (virtualized customer promised equipment), vS-GW (virtualized Serving Gateway), vP-GW (virtualized PDN (Packet Data Network) Gateway), etc., which are carrier network functions.

  The management virtual node (EMS) 32 manages the virtual node (VNF) 31. Specifically, the management virtual node (EMS) 32 manages the network function set in the virtual node (VNF) 31. Each management virtual node (EMS) 32 may manage one virtual node (VNF) 31. Alternatively, one management virtual node (EMS) 32 may manage two or more virtual nodes (VNF) 31.

(NFVI10)
Next, the NFVI 10 will be described in detail.

  The NFVI 10 includes a physical control unit 111, a physical storage unit 112, a physical communication unit 113, a virtualization layer 12, a virtual control unit 13, a virtual storage unit 14, and a virtual communication unit 15. .

  The virtual machine includes a virtual control unit 13, a virtual storage unit 14, and a virtual communication unit 15. Here, the NFVI 10 may be configured to include two or more virtual machines. In the following description, the virtual control unit 13, the virtual storage unit 14, and the virtual communication unit 15 are also simply referred to as “virtual machines” when there is no need to distinguish them. In the following description, a virtual machine is also called a guest machine.

  The physical control unit 111 controls information processing apparatuses (computers) that configure the NFVI 10. For example, the physical control unit 111 according to the present embodiment is configured to include an FPGA.

  The physical storage unit 112 stores information necessary for operating the information processing apparatus constituting the NFVI 10. For example, the physical storage unit 112 is realized using a magnetic disk device, an optical disk device, or a semiconductor memory.

  In particular, the physical storage unit 112 according to the present embodiment includes a shared memory area that is accessible from both the physical control unit 111 and the virtual control unit 13. Details of the shared memory area will be described later.

  The physical communication unit 113 controls processing in which the physical control unit 111 communicates with the virtual machine, NFV MANO2, OSS / BSS4, client 200, and the like. Further, the physical communication unit 113 controls communication processing with a network different from the communication system 100 according to the present embodiment. For example, the physical communication unit 113 is realized using a NIC (Network Interface Card) or the like.

  The virtualization layer 12 executes a process of abstracting hardware. Furthermore, the virtualization layer 12 relays the connection between the hardware, the virtual machine, and the VNF. In other words, the hardware accesses the virtual machine and the VNF via the virtualization layer 12, and executes the process set in the virtual machine and the VNF.

  The virtual control unit 13 executes VNF set in the virtual node 31. The virtual storage unit 14 stores information necessary for executing VNF. The virtual communication unit 15 controls communication processing between the virtual machine and NFV MANO2. The virtual communication unit 15 controls communication processing between the virtual machine and the OSS / BSS 4.

(OSS / BSS4)
The OSS / BSS 4 provides a communication service. For example, the communication carrier manages the OSS / BSS 4.

  Hereinafter, the NFVI 10 will be described in more detail.

  FIG. 3 is a block diagram illustrating an example of modules related to the present embodiment in the internal configuration of the NFVI 10 according to the present embodiment. The NFVI 10 shown in FIG. 3 includes a CPU 120, a main memory (corresponding to the physical storage unit 112 shown in FIG. 2) 130, an FPGA 140, and a memory management unit 150. In the following description, the circuit of the FPGA 140 is referred to as an offload unit (physical control unit 111 shown in FIG. 2) 141. Further, in the following description, a process (process) executed by the CPU 120 is referred to as a non-offload unit (corresponding to the virtual control unit 13 illustrated in FIG. 2) 121. The main memory 130 is also referred to as a first storage unit. The memory management unit 150 is also referred to as a storage area management unit. 3 shows one FPGA 140 and one CPU 120, this is not intended to limit the FPGA 140 and the CPU 120 according to the present embodiment to one, respectively. Similarly, FIG. 3 shows one offload unit 141 and one non-offload unit 121, which includes an offload unit 141 and a non-offload unit 121 according to the present embodiment, respectively. It is not intended to be limited to one. The NFVI 10 may include two or more FPGAs 140, two or more CPUs 120, two or more offload units 141, and two or more non-offload units 121.

  A register (also referred to as a second storage unit) 142 of the FPGA 140 stores an address mapping table 143. The address mapping table 143 stores the correspondence between virtual addresses and physical addresses.

  The main memory 130 includes a shared memory area accessible from both the offload unit 141 and the non-offload unit 121. The shared memory area includes a second storage area and a first storage area.

  The first storage area includes a message buffer (also referred to as a message buffer; mbuf) 132. The message buffer stores packet data.

  The second storage area stores the address (virtual address) of the message buffer 132. For example, the second storage area may store the address (virtual address) of the message buffer using a ring queue type array. In the following description, a case where the address of the message buffer 132 is stored in the second storage area using a ring queue type array will be described as an example. In the following description, a ring queue type array stored in the second storage area is referred to as a ring queue 131.

  The second storage area stores the address (virtual address) of the message buffer 132 using a ring queue type array. As shown in FIG. 3, the second storage area may store a plurality of ring queues (133_1 to 133_n (n is a natural number of 2 or more)).

  The address mapping table 143 stores the correspondence between physical addresses and virtual addresses.

  The NFVI 10 performs packet communication between the offload unit 141 and the non-offload unit 121 using the ring queue 131 and the message buffer 132.

  The packet data transmission source (offload unit 141 or non-offload unit 121) writes the packet data to the message buffer 132. The packet data transmission source (offload unit 141 or non-offload unit 121) stores (enqueues) the address (virtual address) of the message buffer 132 in which the packet data is written in the ring queue 131.

  The packet data transmission destination (offload unit 141 or non-offload unit 121) extracts (dequeues) the address (virtual address) of the message buffer 132 from the ring queue 131. Then, the transmission destination of the packet data refers to the address mapping table 143 and converts the virtual address of the message buffer 132 to the physical address of the message buffer 132. The packet data transmission destination accesses the packet data stored in the message buffer 132 based on the converted physical address.

  For example, it is assumed that packet data is transmitted from the offload unit 141 to the non-offload unit 121. In this case, the offload unit 141 (packet data transmission source) writes the packet data in the message buffer 132. Then, the address of the message buffer 132 in which the packet data is written in the offload unit 141 (packet data transmission source) is stored in the ring queue 131.

  Then, the non-offload unit 121 (packet data transmission destination) takes out the address (virtual address) of the message buffer 132 in which the packet data is written from the ring queue 131. Then, the non-offload unit 121 (packet data transmission destination) refers to the address mapping table 143 and converts the virtual address into a physical address. Then, the non-offload unit 121 (packet data transmission destination) reads the message buffer 132 based on the extracted message buffer 132 address. Then, the non-offload unit 121 (packet data transmission destination) accesses the packet data stored in the message buffer 132 using the converted physical address.

  The memory management unit 150 operates on the OS (Operating System) of the host machine and controls the main memory 130. In particular, the memory management unit 150 manages the ring queue 131 and the message buffer 132.

  Next, the operation of the communication system 100 according to the present embodiment will be described in detail.

  4, 5, and 6 are sequence diagrams illustrating an example of the operation of the communication system 100 according to the present embodiment.

  First, the operation of the communication system 100 according to the present embodiment will be described with reference to FIG.

  First, the client 200 requests the NFV orchestrator 21 to generate a VNF (step S001). Then, the NFV orchestrator 21 receives a VNF generation request (step S002). Here, the client 200 is, for example, a terminal device used by a person in charge of maintenance of the NFV MANO2, an upper device of the NFV MANO2, and the like.

  In step S003, the NFV orchestrator 21 checks whether or not VNF can be configured. Then, if the VNF can be configured, the NFV orchestrator 21 requests the VNF manager 22 to generate the VNF (step S004). Then, the VNF manager 22 receives a VNF generation request (step S005).

  In step S006, the VNF manager 22 requests the NFV orchestrator 21 to allocate resources.

  In step 007, the NFV orchestrator 21 executes resource allocation preprocessing. For example, the NFV orchestrator 21 executes a process for securing resources and reserving allocation as a pre-process for resource allocation.

  In step S008, the NFV orchestrator 21 requests the VIM 23 to allocate resources (control, storage, network) and to set connection between modules.

  In step S <b> 009, the VIM 23 executes assignment of a network that connects modules in the system.

  Next, the operation of the communication system 100 according to the present embodiment will be described with reference to FIG.

  In step S010, the VIM 23 assigns a virtual machine and sets the virtual machine in a network that connects modules in the system. Then, the VIM 23 responds (ACK (Acknowledgement)) to the NFV orchestrator 21 that resource allocation is complete (step S011). Then, the NFV orchestrator 21 responds (ACKs) the resource allocation completion to the VNF manager 22 (step S012).

  The VNF manager 22 notifies the management virtual node (EMS) that the VNF generation has succeeded (step S013).

  In step S014, the management virtual node (EMS) configures the VNF. Specifically, the management virtual node (EMS) sets a specific parameter in an application that configures the virtual node (VNF).

  Next, the operation of the communication system 100 according to the present embodiment will be described with reference to FIG.

  In step S015, the VNF manager 22 notifies the VIM of information on the name of the memory area that stores the ring queue 131 and the message buffer 132. That is, the VNF manager 22 notifies the VIM of information for specifying the second storage area and the first storage area. For example, in the case of Intel DPDK, the host machine and guest machine constituting the NFVI can access data stored in the memory area corresponding to the memory area name by using the memory area name as a search key.

  Note that when executing the process of step S015, the memory management unit 150 has secured the memory areas (second storage area and first storage area) for storing the ring queue 131 and the message buffer 132. It shall be. For example, the memory management unit 150 uses a technique such as IVSHMEM (Inter VM (Virtual Machine) Shared Memory) before executing the VNF generation process, between the host machine and the virtual machine, and between the virtual machines. An accessible shared memory area may be secured using a common virtual address.

  In step S016, the VIM 23 notifies the memory management unit 150 of information on the name of the memory area that stores the ring queue 131 and the message buffer 132. That is, the VIM 23 notifies the memory management unit 150 of information for specifying the second storage area and the first storage area.

  In step S017, the memory management unit 150 creates an address mapping table 143 between physical addresses and virtual addresses. Specifically, the memory management unit 150 creates an address mapping table 143 for converting a virtual address to a physical address.

  In step S018, the memory management unit 150 sets the address mapping table 143 between the physical address and the virtual memory in the FPGA 140. Specifically, the memory management unit 150 sets an address mapping table 143 for converting a virtual address to a physical address in the register 142 of the FPGA 140. Further, the memory management unit sets the virtual address of the ring queue 131 in the register 142 of the FPGA 140.

  In step S 019, the VNF manager 22 responds (ACKs) the end of VNF generation to the NFV orchestrator 21. Then, the NFV orchestrator 21 responds (ACKs) to the client 200 that VNF generation is completed.

  As described above, the NFVI 10 according to the present embodiment includes a shared memory area that is accessible from both the host machine and the virtual machine. The NFVI 10 according to the present embodiment stores packet data in the shared memory area when performing communication between the host machine and the virtual machine. Further, the NFVI 10 according to the present embodiment uses the address mapping table 143 to convert an address between an address (virtual address) on the guest machine and an address (physical address) on the host machine. Therefore, the NFVI 10 according to the present embodiment can realize access to the packet data without executing a process of copying the packet data between the host machine and the guest machine. Therefore, the NFVI 10 according to the present embodiment contributes to improving the communication speed between the host machine and the guest machine.

[Second Embodiment]
Next, a second embodiment will be described in detail with reference to the drawings.

  In this embodiment, the memory management unit creates an address mapping table for converting a virtual address into a physical address. In the description of the present embodiment, the description of the same part as the above embodiment is omitted. Further, in the description of the present embodiment, the same components as those of the above-described embodiment are denoted by the same reference numerals, and the description thereof is omitted. In the description of the present embodiment, the description of the same operational effects as those of the above-described embodiment is also omitted.

The internal configuration of the communication system 100 according to the present embodiment is as shown in FIG. Further, of the internal configuration of the NFVI 10 according to the present embodiment, the modules related to the present embodiment are as shown in FIG.

  7, 8, and 9 are sequence diagrams illustrating an example of the operation of the communication system 100 according to the present embodiment. Since the processing of steps S101 to S109 illustrated in FIG. 7 is the same as the processing of steps S001 to S009 illustrated in FIG. 4, detailed description thereof is omitted.

  Hereinafter, the operation of the communication system 100 according to the present embodiment will be described with reference to FIG.

  When the VIM 23 assigns a network to which the system internal module is connected (step S109 shown in FIG. 7), the virtual machine is assigned and the virtual machine is set in the network to which the system internal module is connected (step S110). ).

  In step S111, the VIM 23 requests the memory management unit 150 to allocate a memory area for storing the ring queue 131 and the message buffer 132.

  In step S <b> 112, the memory management unit 150 creates or allocates a memory area for storing the ring queue 131 and the message buffer 132 in the main memory 130.

  In step S113, the memory management unit 150 creates an address mapping table 143 between physical addresses and virtual addresses. Specifically, as in the process of step S017 shown in FIG. 6, the memory management unit 150 creates an address mapping table 143 for converting a virtual address to a physical address.

  In step S <b> 114, the memory management unit 150 sets the address mapping table 143 between the physical address and the virtual address in the FPGA 140. Specifically, similar to the processing in step S018 shown in FIG. 6, the memory management unit 150 sets the address mapping table 143 for converting the virtual address to the physical address in the register 142 of the FPGA 140. Further, the memory management unit sets the virtual address of the ring queue 131 in the register 142 of the FPGA 140.

  Then, the VIM 23 responds (ACKs) the resource allocation completion to the NFV orchestrator 21 (step S115 shown in FIG. 10). And the communication system 100 which concerns on this embodiment performs the process of step S116-S120 shown in FIG. The processes in steps S116 to S120 shown in FIG. 10 are the same as the processes in steps S011 to S014 shown in FIG. 5 and steps S019 to S020 shown in FIG.

  As described above, in the NFVI 10 according to the present embodiment, the memory management unit 150 creates or assigns the address mapping table 143 and the message buffer 132 for converting addresses. The NFVI 10 according to the present embodiment stores packet data in the shared memory area when performing communication between the host machine and the virtual machine. Further, the NFVI 10 according to the present embodiment uses the address mapping table 143 to convert an address between an address (virtual address) on the guest machine and an address (physical address) on the host machine. Therefore, the NFVI 10 according to the present embodiment can realize access to the packet data without executing a process of copying the packet data between the host machine and the guest machine. Therefore, the NFVI 10 according to the present embodiment contributes to the improvement of the communication speed between the host machine and the guest machine, as in the first embodiment.

  A part or all of the above-described embodiment can be described as in the following supplementary notes, but is not limited thereto.

  (Additional remark 1) It is as the control apparatus which concerns on a said 1st viewpoint.

  (Supplementary note 2) The control according to supplementary note 1, further comprising a second storage unit that stores an address mapping table that associates a physical address with a virtual address, and wherein the physical control unit manages the address mapping table apparatus.

  (Supplementary Note 3) The virtual control unit stores packet data in the first storage area, and stores a virtual address specifying the first storage area in which the packet data is stored in the second storage area. Then, the physical control unit reads the virtual address from the second storage area, refers to the address mapping table, and converts the read virtual address into a physical address that identifies the first storage area. The control device according to appendix 2, wherein packet data is read from the first storage area based on the converted physical address.

  (Supplementary Note 4) The physical control unit stores packet data in the first storage area, and stores a virtual address that identifies the first storage area in which the packet data is stored in the second storage area. Then, the virtual control unit reads the virtual address from the second storage area, refers to the address mapping table, and converts the read virtual address into a physical address that identifies the first storage area. The control device according to attachment 2 or 3, wherein packet data is read from the first storage area based on the converted physical address.

  (Supplementary Note 5) A storage area management unit that stores the address mapping table in the second storage unit is further included, and the storage area management unit includes the first storage area and the first storage area. The control device according to appendix 4, which generates the second storage area.

  (Additional remark 6) It is as the control system which concerns on the said 2nd viewpoint.

  (Supplementary note 7) The control system according to supplementary note 6, wherein the management node generates the first storage area and the second storage area in the first storage unit.

  (Supplementary note 8) The control device stores an address mapping table that associates a physical address with a virtual address, stores a second storage unit, and stores the address mapping table in the second storage unit The storage area management unit further includes an area management unit, and the storage area management unit generates the first storage area and the second storage area in the first storage unit. Control system.

  (Additional remark 9) It is as the control method which concerns on the said 3rd viewpoint.

  (Additional remark 10) It is as the program concerning the said 4th viewpoint.

  It should be noted that the disclosures of the above-mentioned patent documents and non-patent documents are incorporated herein by reference. Within the scope of the entire disclosure (including claims) of the present invention, the embodiment can be changed and adjusted based on the basic technical concept. Further, various combinations or selections of various disclosed elements (including each element of each claim, each element of each embodiment, each element of each drawing, etc.) are possible within the framework of the entire disclosure of the present invention. is there. That is, the present invention of course includes various variations and modifications that could be made by those skilled in the art according to the entire disclosure including the claims and the technical idea. In particular, with respect to the numerical ranges described in this document, any numerical value or small range included in the range should be construed as being specifically described even if there is no specific description.

2 NFV MANO
3 VNF layer 4 OSS / BSS
10 NFVI
11 Hardware resource 12 Virtualization layer 13, 1001 Virtual control unit 14 Virtual storage unit 15 Virtual communication unit 21 VNF Orchestrator 22 VNF manager 23 VIM
31, 31-a to 31-c Virtual node (VNF)
32, 32-a to 32-c Management virtual node (EMS)
100 Communication System 111, 1002 Physical Control Unit 112 Physical Storage Unit 113 Physical Communication Unit 120 CPU
121 Non-offload unit 130 Main memory 131, 133_1 to 133_n Ring queue 132 Message buffer 140 FPGA
141 Offload unit 142 Register 143 Address mapping table 150 Memory management unit 200 Client 1000 Control device 1003 First storage unit 1004 First storage area 1005 Second storage area

Claims (10)

A control device including one or more virtual machines that control a virtual network function and a host machine that controls the virtual machines,
A virtual control unit for controlling the virtual machine;
A physical control unit for controlling the host machine;
A first storage unit accessible by the virtual control unit and the physical control unit;
With
The first storage unit includes a first storage area and a second storage area,
The first storage area stores packet data,
The second storage area stores a virtual address that identifies the first storage area,
The control device specifies the first storage region based on the virtual address stored in the second storage region, and reads packet data from the specified first storage region. A second storage unit for storing an address mapping table that associates a physical address with a virtual address;
The control device according to claim 1, wherein the physical control unit manages the address mapping table. The virtual control unit stores packet data in the first storage area, stores a virtual address specifying the first storage area in which the packet data is stored in the second storage area,
The physical control unit reads the virtual address from the second storage area, refers to the address mapping table, converts the read virtual address to a physical address that identifies the first storage area, and converts The control device according to claim 2, wherein packet data is read from the first storage area based on the physical address. The physical control unit stores packet data in the first storage area, stores a virtual address that identifies the first storage area in which the packet data is stored in the second storage area,
The virtual control unit reads the virtual address from the second storage area, refers to the address mapping table, converts the read virtual address to a physical address that identifies the first storage area, and converts the virtual address The control device according to claim 2 or 3, wherein packet data is read from the first storage area based on the physical address. A storage area management unit for storing the address mapping table in the second storage unit;
The control device according to claim 4, wherein the storage area management unit generates the first storage area and the second storage area in the first storage unit. A control apparatus including one or more virtual machines that control the virtual network function and a host machine that controls the virtual machines;
A management node to which the virtual machine is assigned;
A control system comprising:
The control device includes:
Including one or more virtual machines that control the virtual network function and a host machine that controls the virtual machines;
A virtual control unit for controlling the virtual machine;
A physical control unit for controlling the host machine;
A first storage unit accessible by the virtual control unit and the physical control unit;
With
The first storage unit includes a first storage area and a second storage area,
The first storage area stores packet data,
The second storage area stores a virtual address that identifies the first storage area,
The control system, wherein the control device specifies the first storage area based on the virtual address stored in the second storage area, and reads packet data from the specified first storage area.   The control system according to claim 6, wherein the management node generates the first storage area and the second storage area in the first storage unit. The control device includes:
A second storage unit that stores an address mapping table that associates a physical address with a virtual address;
A storage area management unit for storing the address mapping table in the second storage unit;
Further comprising
The control system according to claim 6 or 7, wherein the storage area management unit generates the first storage area and the second storage area in the first storage unit. One or two or more virtual machines that control the virtual network function and a host machine that controls the virtual machine, a virtual control unit that controls the virtual machine, and a physical control that controls the host machine And a first storage unit accessible by the virtual control unit and the physical control unit, the first storage unit comprising a first storage area and a second storage area A control method for a control device comprising:
Storing packet data in the first storage area;
Storing a virtual address for identifying the first storage area in the second storage area;
Identifying the first storage area based on the virtual address stored in the second storage area;
Reading packet data from the identified first storage area;
Control method. One or two or more virtual machines that control the virtual network function and a host machine that controls the virtual machine, a virtual control unit that controls the virtual machine, and a physical control that controls the host machine And a first storage unit accessible by the virtual control unit and the physical control unit, the first storage unit comprising a first storage area and a second storage area A program that causes a computer that controls a control device including the program to be executed,
A process of storing packet data in the first storage area;
A process of storing a virtual address for identifying the first storage area in the second storage area;
A process of identifying the first storage area based on the virtual address stored in the second storage area;
A process of reading packet data from the identified first storage area;
A program for causing the computer to execute.

Images